Publicación bianual Número 07 // 2010
JUNTA DE ANDALUCÍA. CONSEJERÍA DE CULTURA Conjunto Arqueológico Madinat al-Zahra
Cuadernos de Madinat al-Zahra es una publicación científica bianual editada por el Conjunto Arqueológico de Madinat al-Zahra, que inició su andadura en el año 1987. Recoge trabajos originales que aborden temáticas referentes a la historia y arqueología de al-Andalus y el mundo mediterráneo dentro del marco cronológico de la Edad Media. No obstante, los consejos de redacción y asesor podrán valorar positivamente la inclusión de estudios que den cabida a otros ámbitos y a una ampliación de los límites cronológicos especificados, siempre que contribuyan a la mejor comprensión del periodo. De igual modo, y de forma ocasional, podrán introducirse secciones monográficas o actas de jornadas o reuniones científicas.
DIRECCIÓN ANTONIO VALLEJO TRIANO Conjunto Arqueológico de Madinat al-Zahra CONSEJO DE REDACCIÓN (Miembros de la Comisión Técnica de Madinat al-Zahra) Vocales:
MANUEL ACIÉN ALMANSA Universidad de Málaga CARMEN BARCELÓ TORRES Universidad de Valencia EDUARDO MANZANO MORENO Profesor de investigación del CSIC RUBÍ SANZ GAMO Directora del Museo de Albacete JUAN SERRANO MUÑOZ Arquitecto
CONSEJO ASESOR PATRICE CRESSIER CNRS, Lyon PIERRE GUICHARD Universidad de Lyon II ESTEBAN HERNÁNDEZ BERMEJO Universidad de Córdoba Mª ANTONIA MARTÍNEZ NÚÑEZ Universidad de Málaga ALASTAIR NORTHEDGE Universidad de Paris I VÍCTOR PÉREZ ESCOLANO Universidad de Sevilla
Edita JUNTA DE ANDALUCÍA. Consejería de Cultura © de la edición JUNTA DE ANDALUCÍA. Consejería de Cultura Diseño y maquetación: Carmen Jiménez Diseño de portada: Zum Creativos Imprime: Tecnographic ISSN: 1139-9996 Depósito Legal: SE-8516/2010 Distribución nacional e internacional: 1000 ejemplares
Publicación bianual Número 07 // 2010
ÍNDICE
MISCELÁNEA DE HISTORIA Y CULTURA MATERIAL DE AL-ANDALUS. HOMENAJE A MARYELLE BERTRAND (Textos reunidos por P. CRESSIER, I. MONTILLA TORRES, J. R. SÁNCHEZ VICIANA y A. VALLEJO TRIANO)
05 PRESENTACIÓN Patrice Cressier, Irene Montilla Torres, José Ramón Sánchez Viciana y Antonio Vallejo Triano
06 MARYELLE BERTRAND 06
Maryelle Bertrand (1948-2007) Léon Pressouyre
08
Maryelle Bertrand. Bibliografía 1985-2008
10 LOS SEÑORES DE LA GUERRA 13
Las primeras guerras internas de al-Andalus Eduardo Manzano Moreno
27
Les seigneurs de la Marche (as habu al-tag ri) : les Ban u cAmr u s et les Banu †abri t de Huesca Philippe Sénac
43
Militares en iluminaciones y marfiles: una visión del ejercito califal Juan Zozaya Stabel-Hansen
64 LOS SOPORTES MATERIALES DEL DISCURSO IDEOLÓGICO 67
Le chapiteau, acteur ou figurant du discours architectural califal ? Omeyyades d'al-Andalus et Fatimides d’Ifràà qiya Patrice Cressier
83
Estela funeraria de cronología califal aparecida en Mengíbar (Jaén) María Antonia Martínez Núñez
95
Nuevas evidencias de cecas africanas en época de al-öakam II: al-Mans u rah/al-Mans u riyya y al-Bas ra Alberto Canto García
102 ESPACIOS DE VIDA 105 Excavations in medieval settlements at Volubilis. 2000-2004 Elizabeth Fentress and Hassan Limane 123 Casas y cosas: espacios y funcionalidad en las viviendas emirales del Tolmo de Minateda
(Hellín, Albacete) Sonia Gutiérrez Lloret y Víctor Cañavate Castejón 149 La vivienda tradicional en la cuenca del Mediterráneo: del iw an al qb u’, pasando por el bahw Sakina Missoum 175 Habitat e utensílios na Mértola almóada Susana Gómez, Lígia Rafael e Santiago Macias
196 CASTILLOS Y PALACIOS 199 La fortaleza de Amergo (Marruecos) ¿Otro ejemplo de influencia hispánica en Marruecos? Manuel Acién Almansa 219 Los baños de la tropa de la Alcazaba de Almería: resultados preliminares de la intervención
arqueológica Sophie Gilotte, Ángela Suárez Márquez, Francisca Alcalá Lirio y Francisco Arias de Haro 239 El asentamiento islámico de Giribaile (Jaén). De asentamiento de altura a castillo almohade Juan Carlos Castillo Armenteros, Luis María Gutiérrez Soler y María Victoria Gutiérrez Calderón 263 Los palacios islámicos de Jaén. El palacio de Santo Domingo y los jardines de los Uribe Vicente Salvatierra Cuenca, Mercedes Navarro Pérez y Ángela Esteban Marfil 293 Notes sur les forteresses de la t a ca de Bentomíz (Vélez Málaga) Marie-Christine Delaigue
308 CUEVAS NATURALES, CUEVAS ARTIFICIALES Y OTROS SUBTERRÁNEOS 311 La caverne, refuge de « l'ami de Dieu » : une forme particulière de l'érémitisme au temps
des Almoravides et des Almohades (Maghreb extrême, XIe-XIIIe siècles) Jean-Pierre Van Staëvel 327 Le vocabulaire des grottes et des cavernes dans le Maghreb médiéval à la lumière des sources
arabes Mohamed Meouak 343 Las cuevas de Benaxuay. Un grupo de cuevas-ventana andalusíes en el río Chelva (Valencia) Agustí Ribera 369 Antiguos depósitos de agua en la ciudad de Palma: un patrimonio oculto Maria Antònia Carbonero Gamundí
382 INTERCAMBIOS, HOMBRES Y NATURALEZA 385 Contribución a la historia ambiental de la cuenca del Guadiana Menor (Sureste ibérico):
avances y propuestas de investigación desde la arqueología José Antonio Garrido García 405 Una aproximación a las canteras de piedra calcarenita de Madà nat al-Zahr a’ Antonio Vallejo Triano y Ramón Fernández Barba 421 Comercio mudo / Silent Trade en el Islam Pedro Chalmeta Gendrón 429 1287: onomástica femenina en Menorca islámica Guillem Rosselló Bordoy y Mª Magdalena Riera Frau
434 CRÓNICA DEL CONJUNTO ARQUEOLÓGICO
PRESENTACIÓN En noviembre de 2007, la noticia del fallecimiento de Maryelle Bertrand nos dejó, a todos sus amigos, golpeados y desamparados. Para la mayoría, además, la sorpresa era brutal: con su habitual pudor, Maryelle había callado, durante aquellos fatídicos meses, la gravedad de su enfermedad. De repente, se hacía un inmenso vacío. Todos vivimos entonces un sentimiento de amistad irremediablemente truncada y nos enfrentamos a la añoranza de la complicidad que nos había unido. El vacío no era solo personal, íntimo, sino que era también colectivo y científico. Maryelle había llevado una carrera en cierta forma atípica, parcialmente al margen de las instituciones, pero había participado de pleno en la reflexión que, por aquellos momentos, centraba la atención de los historiadores, en torno a la percepción y a la definición misma de al-Andalus. Estaba presente también en los debates de los primeros años ochenta, en los que se intentaba establecer las reglas de una nueva arqueología que fuera a la vez mejor articulada con el cuestionamiento histórico y más acorde con las necesidades de nuestra sociedad. El tiempo ha mostrado la parte de ilusión que conllevaban tales proyectos, y como se erosionaron frente a la práctica cotidiana que se fue imponiendo. Todavía bajo la emoción causada por su desaparición, y quizá tanto para ayudarnos en nuestro duelo como para recuperar parte de la ilusión pasada, a un grupo de sus amigos nos pareció que convenía rendir un justo tributo a la aportación científica y a la calidez humana de Maryelle. Vicente Salvatierra nos permitió reaccionar en el acto y acogió enseguida una breve semblanza de la vida de Maryelle y su bibliografía completa en la revista Arqueología y territorio medieval1. A más largo plazo, concebimos el proyecto de un homenaje de carácter académico y científico que reuniese contribuciones de los historiadores y arqueólogos de al-Andalus que habían sido los más próximos a Maryelle. Desde el principio, Antonio Vallejo propuso a los Cuadernos de Madànat al-ZahrÄ’ como soporte editorial de este segundo acto. El lector tiene entre las manos el resultado de esta empresa colectiva, asumida por todos con tenacidad y entusiasmo, y a la que –más allá de la diversidad cronológica y de los intereses de cada uno– se ha intentado dar la mayor coherencia temática posible. D. Léon Pressouyre, catedrático emérito de historia del arte medieval de la universidad de Paris 1 – Panthéon Sorbonne, quien dirigió la monumental tesis doctoral de Maryelle y le brindó un apoyo continuado a lo largo de los años, nos aportó, desinteresadamente, su visión personal de la trayectoria profesional y vital de nuestra amiga. Lamentablemente, L. Pressouyre falleció en agosto de 2009, antes de que este homenaje a Maryelle Bertrand haya tomado su forma definitiva2. Ambos, profesor y discípula, quedarán asociados en nuestra memoria. A continuación, las distintas contribuciones vienen agrupadas en apartados sucesivos y complementarios (Los señores de la guerra; Los soportes materiales del discurso ideológico; Espacios de vida; Castillos y palacios; Cuevas naturales, cuevas artificiales y otros subterráneos; Intercambios, hombres y naturaleza). Patrice Cressier, Irene Montilla Torres, José Ramón Sánchez Viciana y Antonio Vallejo Triano
1 “In Memoriam. Maryelle Bertrand (1948-2007)”, Arqueología y territorio medieval, 15, 2008, pp. 9-12. 2 Véase una breve nota necrológica en Bulletin monumental, 2010 (II), pp. 131-132.
CUADERNOS DE MADINAT AL-ZAHRA’ // Nº 07. 2010. P. 5. ISSN: 1139-9996
5
ESPACIOS DE VIDA
EXCAVATIONS IN MEDIEVAL SETTLEMENTS AT VOLUBILIS. 2000-2004 Elizabeth Fentress* and Hassan Limane**
Resumen Se presentan los resultados de la excavación del sector meridional de Volubilis. Estos trabajos, que constituían un programa de cooperación INSAP - UCL, se llevaron a cabo de 2000 a 2005 y fueron dirigidos por los autores de este artículo. Dos zonas se excavaron en extensión. La primera, dentro de la parte del asentamiento defendida por la fortificación del siglo VI, reveló una serie de casas de planta muy sencilla, atribuables a los siglos VII-IX. Su localizacion, sobre más de dos metros de rellenos de destrucción, sugiere una fuerte discontinuidad respecto a la ciudad romana. Extra muros, en la vega del wadi Khomane, un asentamiento integramente nuevo consiste en por lo menos cuatro conjuntos organizados alrededor de patios, de los que uno alberga un ÜammÄm, mientras que otro se componía de dos alas y podría haber constituido un área de recepción. Monedas y cerámicas fechan estos conjuntos desde finales del siglo VIII hasta principios del siglo IX, coetáneos pues de la llegada de Idràs I. Se trataría, por tanto, de la zona de mando y de representación del poder idrisí. Palabras clave: Idràs I, ÜammÄm, baños, Volubilis, excavaciones.
Abstract The paper reports on the INSAP - UCL excavations in the medieval sector of Volubilis between 2000 and 2005, directed by the authors. Two areas were examined by open-area excavation: the first, within the area defended by fortifications of the sixth century, revealed a series of houses of simplified plans, dating from the seventh through the ninth centuries. Their position on top of fully two metres of destruction material suggests a strong discontinuity with the Roman town. Outside the walls, on the flood plain of the Oued Khomane, an entirely new settlement consists of at least four interlinked courtyard structures, one of which encloses a ÜammÄm, while another, with two wings, suggests a reception area. Coins and pottery date this to the late eighth and early ninth century, contemporary with the arrival of Idràs I. It is suggested that the complex constituted his headquarters. Keywords: Idràs I, ÜammÄm, baths, Volubilis, excavations.
*London College. [ elizabeth.fentress@gmail.com ] **INSAP, Rabat. [ hlimane@yahoo.fr ]
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
105
ELIZABETH FENTRESS AND HASSAN LIMANE
The excavations at Volubilis carried out by University College, London and INSAP, Rabat, since 20001 are intended to throw light on the postRoman town, and, particularly, on the occupation of the site under Idràs I, who was declared imam at Walàla in AD 789. Using geophysical survey and traditional excavation techniques we have investigated two main areas, one, site D, inside the Roman town, and one, site B, outside it (fig. 1). The two excavations are in sharp contrast, the one reflecting the housing of the original occupants of the site, the other giving us a glimpse of the urbanism of the later settlers. This brief note should be considered a preliminary report. However, the main lines of the story are by now fairly clear. Very little is known of Volubilis after the departure of the Roman administration in AD 285. The evidence has been ably gathered by A. Akkeraz2. African Red Slip continued to arrive, although in decreasing quantities, until the middle of the fifth century. Coins, too, are rare, although there are a few, as well as some fragments of later Roman metalwork3. It has recently been argued that this limited evidence of external trade implies that the city was under some form of Byzantine control; however there is no obvious link between commercial contact and imperial domination4. In the sixth century a small group of tombstones written in Latin shows the existence of a Christian community that still dated its foundation by the year of the Roman province5. A. Akerraz argues that this community lived in the western third of the city, defended by a new wall running northsouth to link two segments of the earlier Roman walls. The reason for the concentration of settlement in this area may have been the proximity of the Oued Khoumane, whose water would have been necessary if the aqueducts had ceased to function. This area has never been excavated, partially because the medieval town has rendered largely inaccessible the Roman buildings below. Two medieval settlements on the valley floor immediately outside the Roman walls have been excavated since the 1950’s, without, however, giving rise to any significant or, indeed, archival reports.
106
The first half of the eighth century yields an abundant crop of coins, known both from a hoard found just outside the walls and published by D. Eustache6, as well as from the excavations of A. Akerraz to the north-east of the “Maison du Compas”7. For D. Eustache, these indicate the coinage of Abbasid generals, although it has been recently argued that we may be seeing in individuals, such as the otherwise unknown alRaåid b. Kadàm, cAbd al-WaÜab and MuÜammad b. õalifa leaders of an independent city state8. Many others simply bear the name of Walàla. The texts referring to the arrival of Idràs I in 788 are, however, categorical. The town was said to be in the control of the AwrÄba tribe, who welcomed the descendent of cAlà with open arms, and declared him imam shortly thereafter9. Within three years he had consolidated his hold on much of the area, founded the first settlement at Fez10, and proceeded to mint coins11. He died in 781, leaving a pregnant AwrÄba wife and his faithful slave, Raåid, who acted as regent until the majority of Idràs II. At this point the court departed for Fez, leaving the AwrÄba in control of the town.
1. EXCAVATIONS INSIDE THE ROMAN WALLS: SITE D12 The first excavation site was selected for a number of reasons. First, it lay near the tourist circuit of the upper town, and we felt that it might provide an opportunity to present the later occupation of Volubilis as part of the regular circuit. Second, it showed the remains of two relatively intact buildings. Third, it lay 20 m from the late city wall, near a possible gate. A modern track runs from this point down to the wÄdà, and may follow an older route to the southwest gate in the Roman wall. However, the site is peripheral to the center of the late Roman town, which covers 18 ha. It may reflect a less dense occupation, and public buildings are hardly to be expected. During the first two seasons domestic buildings were excavated at either end of the site, revealing unexpected traces of a late-nineteenth century
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
EXCAVATIONS IN MEDIEVAL SETTLEMENTS AT VOLUBILIS. 2000-2004
occupation. This prompted the use of Accelerated Mass Spectrometry (= AMS) to reinforce dating from pottery and coins: all three techniques are still in use, but the dates should still be regarded as provisional. Fig. 2 gives a schematic idea of the building sequence. The earliest, and as yet unexcavated, phase of the site’s stratigraphy that we have been able to observe is a massive destruction layer composed of falls of mud brick and stone masonry at least 1,5 m deep. Within this, later cuts and pits reveal standing walls, but no fuller plans are yet available. The date of this destruction remains obscure, although a calibrated AMS date of 327-412 from the top of the destruction sequence seems to support our view that it took place at the end of the fourth century or the beginning of the fifth. Our earliest calibrated AMS date from the upper occupation is AD 650-879, which suggests that the reoccupation of the site in this area dates to the seventh century, somewhat later than the construction of the western rampart, which A. Akerraz suggests was built in the sixth century. The idea of an abandonment in the fifth century ties in with the lack of coin or pottery finds from the second half of the
fifth century. A small sondage showed that the depth of the destruction of the Roman buildings was over 1,5 m: while the skeleton of a cat and an amphora crushed in situ suggest that it was sudden. It is not impossible that a major earthquake could account for the abandonment of the site. The destruction of all the buildings on the site would have left an irregular surface with abundant clay from the mud bricks, and large blocks of building material. The first activities visible are thus the creation of flat spaces on which to build, and the digging of pits to recover earth and stones for new buildings13. The earliest of these buildings is found in the lower half of the site, where building I is partially preserved (pl. 1). To the north is a long room, F, measuring 3,5 x 9 m and paved with a succession of hard white plaster floors. The walls, too, were plastered. There is no certainty as to the position of the door, as the walls have been partially eroded away. However, a space to the south of the room, room K, may have been partially covered with a lean-to roof. This space may have housed a stable or, perhaps, a workshop, as the large stone vat containing lime and the carefully
Plate 1. Site D, building I.
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
107
ELIZABETH FENTRESS AND HASSAN LIMANE
posed stone in the corner seems to suggest. To the north of the building were cut large pits filled with ash mixed with bone, pottery and a few ceramic wasters. The fills strongly suggest ceramic or glass production on the site, although it is unlikely that the pits were themselves kilns, as the earth into which they were cut does not appear burned. The stratigraphic links of a second building, II, to building I are not entirely clear, but it may be assumed to be of roughly the same period (pl. 2).
Here, the initial terracing clearly revealed earlier buildings, for two of the walls were built along walls founded much deeper. The other walls were roughly built, without foundations and reinforced at intervals with wooden posts. The floor of the room was cut down slightly into the surrounding clay. Measuring 8,5 x 4,2 m., the plan of the building is relatively clear, although, again, the position of the door is uncertain: it was probably to the south. The southern 2/3 of the space was floored in beaten-earth, with a hearth and a
Plate 2. Site D, building II.
108
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
EXCAVATIONS IN MEDIEVAL SETTLEMENTS AT VOLUBILIS. 2000-2004
domestic silo to the west. A dividing wall separates this space from a room with a very irregular floor 30 cm. below the level of the main room. Traces of burning are found in one corner, but in general this space is so rough as to suggest that it was used for storage or animals. A large post-hole occupies the center of this space. The general arrangement of the whole building strongly suggests the individual units studied in the Kabylie by P. Bourdieu14: the posthole would have held the main roof beam, as well as providing support for a loft above the stable/work space (fig. 3). The only notable difference is the orientation of the building, with the short end facing the south rather than the east. Outside the structure to the west were three cylindrical silos, while the area to the east was badly cut by large, irregular pits that may have been used to extract earth for the construction of the upper walls. A rough wall seems to have separated this property from that of building I, to the north, although its line is now only indicated by a few blocks.
A new structure, building III, was built next to building I in a second phase, again involving the terracing of the hillside (pl. 3). Measuring 10 x 5,2 m, its construction is very typical of the Roman walls of the site, with re-used orthostats marking the door and regular, coursed stone masonry up to a height of 1 m. Over this we must imagine that the upper parts of the walls were built in pisĂŠ de terre or mud brick and that the ceiling was of perishable materials, such as reeds on a wooden frame covered with earth, as there were no stones or tiles in the destruction material covering the floor. One door opened onto an empty space to the east, flanked by the north wall of building I, while another opened onto more irregular terrain uphill to the north. This may have been used for stabling. In a second phase, the large room was divided down the middle by an irregular stone wall, and the western room was paved with flat stones, while the eastern room was covered with a plaster of pink earth mixed with a little lime. We may suggest that this was a functional division
Plate 3. Site D, building III.
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
109
ELIZABETH FENTRESS AND HASSAN LIMANE
between domestic space and stabling or storage: the former hypothesis is based on the use of paving stones, which might be assumed to be healthier for the animals than beaten earth, and easier to clean. In a subsequent phase all three buildings were replaced by similar units further up the hill. III was replaced by VI, which re-used its east wall, rebuilding in crude masonry based on large spoliated blocks. I was replaced by IV, of roughly the same size, and also floored with white plaster. The main room, E, was entered from the south, and a hearth was found against its south wall, as well as a domestic silo almost 1,6 m deep in the southeast corner of the room. Outside we may again suggest that room H was used for storage or for stabling, while a large cylindrical silo to the southeast may have belonged to the same property. Finally, II was replaced by V, partially excavated to the east of the site. To the south of room I the modern track covers a road running downhill, with carefully constructed drainage channels on either side. From the surface of this road a silver dihram of Idràs I suggests a possible date for its construction, while an AMS date ranging from AD 818 to 902 from its floor gives a general idea of the date of this occupation. The site thus reveals what seem to be three individual properties, each composed of a single rectangular building, subdivided in two cases into
Plate 4. Spolia from the Arch of Caracalla reused in the bath.
110
domestic space and stabling or workrooms. The central property seems to have had an annexe in both phases, and may have been connected to pottery or glass production. There is no trace of violence in the abandonment of the three properties, but the general lack of tenth century pottery suggests that by that time the town had retracted from this more peripheral area.
2. SITE B Excavation at site B15 was carried out in order to understand the surroundings of the small bath complex first excavated by B. Rosenberger in the 1960’s. This excavation has left no documentation, but sondages in the complex carried out by A. El Khayari in 1993 recovered three Islamic coins, of which one apparently dates to the reign of Idràs I16. In 2000, a magnetometry survey was carried out by K. Brown, showing a series of regular alignments, particularly intriguing as their orientation coincided with that of all the early mosques in Morocco17. Excavation began with the cleaning of the large site excavated by B. Rosenberger, which took an entire season, as flooding and erosion had left substantial deposits of earth over the site. In order to recover some intact stratigraphy the excavation was extended to the south, with the substantial alluvial overburden removed by machines. A smaller trench was excavated through B. Rosenberger’s balk to connect the excavated area to the Roman city wall to the west, with the aim of excavating an undisturbed sequence through the whole occupation. The earliest building on the site, with the exception of the rampart and a kiln dating to the first century, is the bath complex (fig. 4). This comprises four rooms set at right angles to one another. The building consisted, initially, of an undressing area, or “bench room” (maslah or maslaá, in medieval terminology),18 with a cold plunge bath with three steps leading onto it. In a second phase, an entry vestibule was added to the cold room, creating a bent entrance to protect bathers from a direct view from outside. At the other end of the bench room a door led to a
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
EXCAVATIONS IN MEDIEVAL SETTLEMENTS AT VOLUBILIS. 2000-2004
vestibule or winter changing room (waséÄnà barrÄnà), where water fed from the cistern to the south could be used for preliminary washing, and an opening to a drain might have allowed for its use as a latrine. The room was decorated with a large shield in relief carefully set off by red plaster (pl. 4): this is a block spoliated from the Arch of Caracalla. At a roughly right angle to this room a low door led to a warm room (waséÄnà ÑuwwÄnà), commonly used for depilation and a hot room (ÑuwwÄnà). The final two rooms were both covered with a barrel vault. The hot room was heated by a furnace located at its eastern end, whose heat passed under the floor through an axial channel with four smaller ducts at right angles to it (pl. 5). Draught was probably assured by chimneys against the walls at the ends of the westernmost channels. The furnace also heated water in a masonry basin placed on top of it, outside the hot room. A circular hole in its floor suggests that, as at Zenobia and Qasr el Hayr East, the fire was in direct contact with a copper plate, either a circular disk or a cauldron set inside the basin.19 Water from
Plate 5. Underfloor heating duct in hot room.
this basin was manually fed into the two small tanks in the hot room through pipes. It does not appear that the steam from the hot water was fed into the hot room, largely because the space in which the furnace and basin were found does not seem to have been sufficiently enclosed.20 To the west of the baths, towards the wÄdà, no further buildings were recovered, although a stony compacted surface may indicate a road. This seems to have been the western limit of the site, as the absence of any buildings or stratigraphy in a small sondage carried out to the southwest in 2001 confirmed. To the north, lay an irregular building, enclosed between the two wings of the baths and the rampart, and reached by a gate next to the baths, flanked by two massive orthostats (fig. 5). In the center of the building was a large courtyard where a large number of silos was found. At least ten of these were substantial grain silos, cylindrical in form and as much as 2 m in diameter, with depths of up to 2 m. Only in a few cases was any treatment of the sides observed, although in one case a clay coating filled in the spaces between the gravels of the underlying alluvial deposits. Along the west side of the baths there were smaller pits, two of which contained human skulls, and one the articulated body of a dog: no satisfactory explanation has been found for these. The larger pits, however, appear to follow relatively regular alignments, parallel to the wings of the building, and in general do not cut each other. This suggests that they were roughly contemporary. The easiest explanation for this grouping is that they formed part of an agadir, or area of collective grain storage.21 However, traces of productive activity, such as a large piece of iron slag, are also present, and we may perhaps see the structure as serving both productive and storage functions. The area to the south of the baths is very different. Here a series of orthogonal walls delimit at least three separate properties, occupied by large buildings characterized by courtyards. The first of these, building I, was excavated over the period 20012003, while the other two were revealed during the 2004 season. Building I abuts the bath complex (pl. 6). It comprises two wings, separated by a wide cobbled courtyard: over all, it measures 26 x 20 m.
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
111
ELIZABETH FENTRESS AND HASSAN LIMANE
Plate 6. Site D. Building I, from the west.
The two wings are both well-built, with walls in a mixture of orthostats and coursed rubble masonry: their upper sectors were presumably in pisĂŠ de terre. The excavated area of the eastern wing comprises two rooms each partially paved with irregular stone slabs set in a beaten-earth surface. The western room opened directly onto a cobbled street to the east, with another door leading onto the courtyard. Here the paving consisted of a series of plastered floors, of which the lowest was a fine pink clay mixed with earth and lime. Against the south wall a section of this floor about a meter wide was raised 20 cm. It is not clear whether this division was the result of the deliberate creation of a bench, or whether the presence of cushions or rugs at this end impeded the wear that took place elsewhere in the room: in either case this area would have been used for seating. On the walls were traces of a fine grey lime plaster painted red, similar to the red paint in the hot room of the baths.
112
The courtyard was reached by a large door marked by two orthostats at the southeast corner of the building. This door is angled at 45% to the other walls of the building, and led in from a large space, as yet unexcavated, that may represent a sort of forecourt, or area in which livestock was kept. The south wall of the courtyard does not present any visible doors, so the range of rooms to the south of it must belong to another property, building II, although there is only a single party wall between them. This building had at least two ranges of rooms, opening, again, onto a single courtyard. These were 2 m wide, on the north and west sides of the court. All rooms are paved with beaten earth, and one can be clearly identified as a kitchen. A further property, building VI, apparently distinct from the first two, lies in the south-east corner of the excavation. A courtyard apparently occupied its northern sector, although nothing is visible of this apart from a hearth built against the west wall.
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
EXCAVATIONS IN MEDIEVAL SETTLEMENTS AT VOLUBILIS. 2000-2004
All of these buildings went through considerable changes in the course of their occupation. In the case of the first building, the eastern wing was subdivided at some point, and repaved with a sequence of three floors. The courtyard seems to have undergone a rapid accumulation of earth between two phases of cobbling, possibly due to alluvial deposition. At a later stage a one-roomed structure was built at the north end of the courtyard, free standing but acting, in effect, as a northern wing. The southern house was abandoned and destroyed; its place was taken later by a new wing along roughly the same plan, with two long, adjoining rooms with doors opening to the south. A posthole in the center of one of these suggests that the roof was partially sustained by posts. To the north of the baths the collective granary went out of use, and was replaced by a number of houses, apparently composed of one or two rooms. We have a very patchy view of this phase, due to the damage caused by the earlier excavations. However, the house excavated near the rampart gives us a good view of the plan of these buildings: again, two substantial rooms are placed end to end, with a door on the north side. In the case of this building the east wall apparently abutted the rampart itself, at this point partially destroyed, with its facing removed and only the rubble core still standing. This house used partial paving on its floors, set into the usual pinkish plaster. Outside it lay a horseshoe-shaped basin, lined with hydraulic mortar. Traces of two other such basins were found on the site. Their explanation is not certain: possibly they were used for washing, although they might be interpreted as watering troughs for animals. Some of the smaller pits may belong to this phase, but, in general, the larger silos are sealed by walls or layers associated with the domestic use of the site. Finally, the bath building, too, seem to have been abandoned, as later walls cross it at two points. A thin layer of yellow clay seems to separate the use of the baths from the domestic occupation of the site. This may have been due to disastrous flooding: in fact, the orthostats of the west wall of the frigidarium, the wall nearest to the Oued Khomane, are
sharply tilted to the north, suggesting the violent thrust caused by the torrent. The phases of the site both to the north and to the south of the baths seem to be relatively consistent, and the broad outlines of the occupation are clear. The earlier structures consist of at least four substantial, interlocking courtyard buildings: a collective granary, a bath complex and, to the south, three very substantial buildings with large courtyards and orthogonal plans. That closest to the baths may have been a public building, as its door, opening directly onto the street, and the high quality of its floors and interior walls seems to suggest. The huge courtyard points in the same direction. It may indeed be a very impoverished version of the dÄr al-imara, or governor’s palace. This would suggest that the mosque, if it existed, is to be found between this building and the rampart, perhaps south of the courtyard of building VI. Building II, with its kitchen and simple rooms, should probably be interpreted as a domestic structure. In a subsequent phase, separated, perhaps, by a period of alluvial deposition, all of these structures were replaced by relatively poorer buildings, of a domestic nature. Our knowledge of the chronology of these phases of settlement remains to be confirmed by the study of the pottery and the coins22. Most of the coins are pre-Idrisid, but their use could have continued well into the ninth century. Those found by A. El Khayari under the baths are still under examination: as this is certainly the earliest building on the site its date is key for the interpretation of the rest. The pottery from the earliest layers associated with the courtyard south of the baths is consistent with a late eighth century date. A radiocarbon date from a later hearth north of the baths gives AD 763 AD to 829. Our tentative conclusion is that the occupation of this site dates from the time of Idràs I. An interruption, perhaps due to flooding may have occurred some time in the ninth century. The site was reoccupied some time later by a series of houses, which remained in occupation until the end of the tenth century. All this is still very preliminary, and awaits confirmation from the further study of the pottery and coins.
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
113
ELIZABETH FENTRESS AND HASSAN LIMANE
3. DISCUSSION The two sites are very different, despite their similar chronologies. In the one case we can see a series of simple houses, built in masonry familiar from the Roman buildings on the site, but whose plans are reduced to one or at most two rooms. Provisions for stabling in two of these suggest that animals were a kept within them, while the relatively undefined properties around them suggest the rural atmosphere of a village more than that of a town. That said, at 16 ha the RomanoBerber settlement at Walàla was as substantial as many better-known towns of the eighth century, while the sixth century town walls certainly suggest collective organization for defense. The single rooms remind us of many Berber domestic arrangements, in the Kabylie and elsewhere23. There seems little doubt that site D gives us a glimpse of a post-Roman Berber community, whose commerce with the outside world was much reduced but not entirely absent, as a glazed jug apparently imported from Spain in the ninth century seems to show24. The settlement on the valley floor is of quite another type (fig. 6). First, the bath building, if small, is strikingly well executed. The drains and water adduction, the heating system, the hydraulic mortar of its cold pool, even the construction of the vaults, are all strongly reminiscent of Roman techniques25. However, certain aspects of the building seems to refer to more eastern models. The provision of benches around the walls of the changing room is not paralleled in late Roman bath architecture, and the use of holes in the vaults for lighting, as we can see in the roof of the warm room, also predicts the elaborate pierced vaults of later baths. But it is the system used for heating the baths that is most clearly relates to Islamic buildings in Syria and even further east. The room was heated by a fire in the praefurnium, fed from the back of the building in a room with no connection to the hot room. As we have seen, the use of a copper disk or cauldron above the praefurnium in the hot water basin pierced for that purpose has parallels only at Zenobia and Qasr el Hayr East. But even more striking is the use of a main chan-
114
nel with lateral ducts, rather than a hypocaust floor, to provide underfloor heating in the hot room, with some heat provided by the chimneys which drew the hot air and smoke upwards in the corners of the room. The system is not used in the Umayyad baths of Syria, which all have floors supported on pillars in the Roman technique. Indeed, it can only be paralleled in the eleventh century baths of Lashkari Bazaar, a Ghaznavid palace in Afghanistan.26 However, a rather similar technique appears to have been used at SÄmarrÄ, although there the floor was placed on low arcading rather than isolated channels.27 The wide channel running down the center of the floor may be seen as a precursor of those found in Ottoman baths, where the Roman hypocaust was generally abandoned in favour of the new technique.28 In later baths the distinction between the floor over the main duct and the rest of the floor is made clear by the term ‘fire slab’, or bilat al-nar.29 The technique was clearly perceived to be eastern, or Persian, by the tenth century author alMuqaddisà, whose comment on a ÜammÄm in Jerusalem is quoted by Martin Dow “You cannot remain in the inner room of the ÜammÄm (in Faris) because of the heat. I heard some of the servants of my late father say Abë al-FaraÑ al-†àrÄzà made a mistake in the ÜammÄm which he built by the gates of al-Asbat because he introduced fire under some of the inner room. And this was not the case, but rather he saw that the designs of †am in this respect were different from the designs of Faris, and made some of the room on the designs of his own land and the rest of it on the designs of †am”.30 The heating system of the ÜammÄm at Volubilis was clearly built “from the designs of Faris”, as, indeed, the Samarran parallel suggests. If the builders of the baths were almost certainly local, technical suggestions were being made by someone well-acquainted with the baths of Baghdad and SÄmarrÄ. The building is clearly public, opening onto what was an open space at the time of its construction. In time, this space seems to have been walled off and devoted to a collective granary, with rooms around it: again we can still see an evident public
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
EXCAVATIONS IN MEDIEVAL SETTLEMENTS AT VOLUBILIS. 2000-2004
use. The situation is less clear in the case of the building to the south of the baths, but even here the huge court and formal decoration of the east wing suggest a public context. Only the southernmost courtyard building, with its kitchen and irregular rooms, has a clearly domestic destination. The whole quarter was clearly laid out according to an orthogonal scheme, in contrast to the haphazard, organic growth apparent in site D. That this layout was aligned on the same orientation as the majority of the pre-twelfth century mosques in Morocco is perhaps not a chance, although we still have no clear idea as to the position of the mosque itself, or even certainly that it existed. We must see here at least some traces of an orderly, planned settlement, laid out perhaps by Idràs I or his followers, to serve as a headquarters.31 After a possibly short interruption marked by alluvial deposition, the lower site was reoccupied for a brief period in the later ninth or tenth century, although we have no evidence for the truth of the comment of al-Bakrà that there were Rabedis on the site at that time.32 This reoccupation on site B, shows houses arranged with far less order than in the preceding period. Nor do they fully conform to the more “Arab” type seen in the first buildings on site B, in that we have no evidence for courtyard buildings, and the proportions of the individual buildings resemble more closely the one and tworoomed houses already present on site D. We seem, then, to be seeing a hybrid type in this period, containing elements of both of the previous types of house: the single-celled structure has been modified by the addition of a courtyard, with the possibility of building original rooms opening onto the same space.
4. AN OVERVIEW OF THE POST-ROMAN SETTLEMENT One major section of the city has been left out of this treatment. This is the extra-mural settlement to the north of site B, enclosed by a second meander of the web. Traditionally known at Volubilis as the “Quartier Arab”, the site served as a mine for
Roman inscriptions, and was serially excavated by a significant number of archaeologists. D. Eustache publishes the only known plan of the area, which is highly schematic, to say the least33. From it, however, we can reach two conclusions. First, two hoards and one buried gold coin were discovered here. The hoard studied by D. Eustache contained coins dating later than AD 742. This would tend to demonstrate that the area was occupied in AD 742, and that it was very likely destroyed near that date. One hoard may be the result of chance, but the three separate finds seem too many for coincidence. A second point is that at the center of the site is a tower, still visible today, in which two bodies were found, one central and one to the side. The tower is clearly one of the earliest features in the area, as an examination of the standing remains show, but its rough construction and heavy use of spolia argue against the interpretation of the structure as a Roman mausoleum. We might suggest that it was built as a tower for funerary or defensive purposes (it lay at the center of the little settlement, and right outside the gate of the Roman city) and that it was later used, or reused, for burial. These two points suggest a very tentative hypothesis. First, that the extramural settlement in this area was of Abbasid date, and came to an end during one of the Berber revolts of the midcentury.34 Second, the tower might have been used for burial at a later date, perhaps that of a distinguished individual. Be that as it may, the pattern of new Arab settlement at Walàla is clear. In both cases, the ?Abbasid and the ?Idrisid settlements were extramural, next to the earlier town but not part of it. This is a pattern already set in the eastern Mediterranean at sites like Amman35 and Ayla36, and apparently followed by the first Islamic settlement at Tlemcen/Agadir37. At Tlemcen, the congregational mosque was built straddling the late-Roman city wall of the town of Pomaria, acting as the interface between the two communities. At Walàla, if we are not yet sure of the position of the mosque, the two communities are at least coming into sharper focus.
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
115
ELIZABETH FENTRESS AND HASSAN LIMANE
Islamic Roman Perimeter Wall Excavations
Fig. 1. Plan of Volubilis, with visible medieval remains in black.
116
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
EXCAVATIONS IN MEDIEVAL SETTLEMENTS AT VOLUBILIS. 2000-2004
Fig. 2. Site D, phase plans.
Fig. 3. Site D. Building II, sketch reconstruction (Fernanda Palmieri).
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
117
ELIZABETH FENTRESS AND HASSAN LIMANE
Fig. 4. Site B, plan of the ÜammÄm (Mohammed Alilou, Fernanda Palmieri).
118
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
EXCAVATIONS IN MEDIEVAL SETTLEMENTS AT VOLUBILIS. 2000-2004
Fig. 5. Plan of site D, eighth century phase.
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
119
ELIZABETH FENTRESS AND HASSAN LIMANE
Fig. 6. The eighth century complex, sketch reconstruction (Fernanda Palmieri).
120
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
EXCAVATIONS IN MEDIEVAL SETTLEMENTS AT VOLUBILIS. 2000-2004
BIBLIOGRAPHY AKERRAZ, A. (1996): Le Maroc du Sud de Dioclétien aux Idrissides, Unpublished doctoral thesis, Paris. AKERRAZ, A. (1985): “Note sur l'enceinte tardive de Volubilis”, Bulletin Archéologique du Comité des Travaux Historiques, 19 B, pp. 429-436. AKERRAZ, A. (1998): “Recherches sur les niveaux islamique de Volubilis”, in P. CRESSIER and M. GARCÍA-ARENAL (eds.), Genèse de la ville islamique en al-Andalus et au Maghreb occidental, Casa de Velázquez-CSIC, Madrid, pp. 295-304.
AL-BAKR∞ (1965): Description de l’Afrique
Septentrionale, ed. and tr. M. G. DE SLANE, Paris.
BONINE, M. (1991): “Islamic cities in Morocco”, Muqarnas, 7, pp. 50-72. BOURDIEU, P. (1990 -reprinted-): “The Kabyle House or the World Reversed”, in The Logic of Practice, Paris, pp. 271-283. BRETT, M. and FENTRESS, E. (1996): The Berbers, Oxford. BRETT, M. and FENTRESS, E. (2000): “Social Relations and Domestic Space in the Maghreb”, in A. BAZZANA and É. HUBERT (eds.), Castrum VI. L'espace domestique dans le monde méditerranéen au Moyen Âge, RomeMadrid, pp. 15-26. CHARPENTIER, G. (1995): “Les petits bains protobyzantins de la Syrie du Nord”, Topoi, 5, pp. 219-247. CHIKAOUI, N. (2002): “Savoir-faire technique dans l’architecture de terre au Maroc (Sud, préRif)”, Bulletin d’Archéologie Marocaine, XIX, pp. 393-422.
EL-KHAYARI, A. (1994): “Les thermes extra muros à Volubilis”, L'Africa Romana, X, Sassari, pp. 301-312. EUSTACHE, D. (1956): “Monnaies musulmanes trouvées à Volubilis”, Hespéris, XLIII pp. 133-195. HOLOD-TRETIAK, R. (1970): “The Bath”, in O. GRABAR, City in the Desert, Qasr el Hayr East, Cambridge, Mass, pp. 90-97. LAUFFRAY, J. (1991): “La palestre et les bains”, Halabiyya-Zenobia, place forte du limes oriental et Haute-Mésopotamie au VIème siècle, t. II, Paris, pp. 113-129. LENOIR, E. (1983-1984): “Volubilis des Baquates aux Rabedis: une histoire sans paroles?”, Bulletin d’Archéologie Marocaine, XV, pp. 299-348. LÉVI-PROVENÇAL, É. (1938): “La fondation de Fès”, Annales de l’Instiut d’études orientales, 4, pp. 23-53. NORTHEDGE, A. (1992): Studies on Roman and Islamic Amman. I: History, Site and Architecture, Oxford. PAPI, E. (2002): “Mauretania Tingitana in epoca tardo-antica: Great Expectations”, Journal of Roman Archaeology, 15, pp. 703-706. PAUTY, E. (1944): “Vu d’ensemble sur les hammams de Rabat-Salé”, Revue africaine, 88, pp. 202-226. REDMAN, Ch. L. (1983): “Comparative Urbanism in the Islamic Far West”, World Archaeology, 14, pp. 355-377. ROSENBERGER, B. (1964-1965): “Une mine d’argent au Moyen Âge marocain”, HespérisTamuda, V-VI, pp. 15-77. ROSENBERGER, B. (1998): “Les premières villes islamiques du Maroc: géographie et fonctions”, in P. CRESSIER, and M. GARCÍAARENAL (eds.), Genèse de la ville islamique en al-Andalus et au Maghreb occidental, Madrid, Casa de Velázquez-CSIC, pp. 229-256.
DAHMANI, S. (1983): “Note sur un exemple de permanence de l’habitat et de l’urbanisme de l’époque antique à l’époque Musulmane: Agadir-Tlemcen”, Bulletin Archéologique du Comité des Travaux Historiques n.s.,19 B, pp. 437-337.
SCHLUMBERGER, D. (1952): “Le Palais Ghaznevide de Lashkari Bazar”, Syria, 29, pp. 251-270.
DOW, M. (1996): The Islamic Baths of Palestine, Oxford.
THÉBERT, Y. (2003): Thermes Romains d’Afrique du Nord et leur contexte méditerranéen, Rome.
ÉCOCHARD, M. and LE COEUR, C. (1942): Les Bains de Damas, Beirut.
VILLAVERDE VEGA, N. (2001): Tingitana en la Antigüedad Tardia (Siglos III-VIII), Madrid.
EL-HARRIF, F. Z. (1998): “Monnaies islamiques trouvées à Volubilis: liberté locale et pouvoir abbaside”, Actes des Ières journées nationales d'Archéologie et du patrimoine. Rabat 1-4 juillet 1998, vol. 3, Rabat, p. 142-159.
WHITCOMB, D. (1994): “The Msir of Ayla: Settlement at al-‘Aqaba in the Early Islamic Period”, in G.R.D. and A. CAMERON eds., The Byzantine and early Islamic Near East II, Princeton, pp. 155-170.
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA
121
ELIZABETH FENTRESS AND HASSAN LIMANE
Notas 1 The excavations, sponsored by University College London and the Institut National des Sciences de l’Archéologie et du Patrimoine, are directed by ourselves, and form part of a larger project that comprises the preparation of a management plan, under the direction of Gaetano Palumbo of UCL and the World Monuments Foundation and the conservation of significant buildings. Sponsorship has come from the British Museum, the World Monuments Foundation, the McAslan Foundation, and the Volubilis Foundation, to whose president, Rita Bennis, we are particularly grateful for her tireless energy and support. 2 AKERRAZ 1985; ID. 1996; ID. 1998; see also LENOIR 1983-1984; VILLAVERDE VEGA 2001. 3 VILLAVERDE VEGA 2001, pp. 157-174. There are 29 coins for the fourth century, and one each in the fith, sixth and seventh centuries (p. 171). This hardly echoes a vibrant urban economy. 4 VILLAVERDE VEGA 2001; see the review of E. PAPI (2002). 5 VILLAVERDE VEGA 2001, pp. 405-407, with previous bibliography. 6 EUSTACHE 1956. 7 The coins from this excavation are published by F. Z. ELHARRIF (1998), who is also responsible for the study of our own coins.
CHARPENTIER 1995, p. 231. For the masonry basin at the sixth century baths at Halabiyya-Zenobia: LAUFFRAY 1991, p. 125; for Qasr el Hayr East: HOLOD-TRETIAK 1990, pl. 175: the hole is not mentioned in the text, but is clearly visible on the photograph. 20 This point is made by THÉBERT 2003, p. 422. 21 Such silos were in use in the area until recent times: today, in the nearby village of Ferdassa, an area now used for housing is known to be “the area for silos”. 22 The pottery is being studied by Victoria Amoros Ruiz and Abdallah Fili and the coins by Fatima-Zohra El-Harrif. The following remarks are based on their preliminary conclusions. 23 On this: BRETT and FENTRESS 1996; FENTRESS 2000. 24 The sherd, a “jarrito tipo pequeño”, was identified by Victoria Amoros Ruiz. 25 The point is made by el EL-KHAYARI 1994. 26 SCHLUMBERGER 1952. 27 A. Northedge, pers. comm. 28 ECOCHARD and LE CŒUR 1942, p. 25-29 and fig. XI on the baths of Damascus; for Rabat: PAUTY 1944. 29 DOW 1996, p. 3. 30 Ibid. p. 26.
8 IBID.
31 Clear differences in the food consumed at the two sites, as
9 Both EUSTACHE (1956) and B. ROSENBERGER (1998), provide
well as in the pottery, will be detailed in the environmental studies under the general direction of Dorian Fuller, and the pottery studies carried out by Victoria Amoros-Ruiz and Abdallah Fili. For a similar approach see C.L. Redman (1983).
useful summaries of the evidence. 10 LÉVI-PROVENÇAL 1938. 11 The suggestion is that of B. ROSENBERGER (1964-65 and 1998). 12 Excavation at site D was carried out under the supervision of Ali Ait Kaci. 13 The digging of pits to recover earth is general even today: see N. CHIKAOUI 2002. 14 BOURDIEU 1990.
32 AL-BAKR∞ 1965, p. 295: however, the identification of Volubilis/Waliali with the Walàli of this passage is not entirely convincing, as it occurs in a section on the road between Fez and Kairouan, which is unlikely to unlikely to have run westwards or through the Zerhoun. 33 EUSTACHE 1956, p. 137. 34 Here we must note the objections of Aomar Akerraz, who
16 EL-KHAYARI 1994.
believes that the extension to the Roman city wall which encloses this settlement to the north is of the same date as that of the sixth century wall which forms the eastern boundary of the post-Roman settlement. This question would have to be resolved by stratigraphic excavation in the area.
17 BONINE 1991.
35 NORTHEDGE 1992.
15 The three main areas of excavation in site B were supervised by Helen Dawson, Guy Hunt and Tarik Moujoud.
18 R. HOLOD-TRETIAK 1970, p. 95, argues that the medieval terminology is more appropriate than a Roman one for describing the standard rooms in a ÜammÄm.
122
19 The suggestion that the hole held a copper disk is made by
36 WHITCOMB 1994. 37 DAHMANI 1983.
CUADERNOS DE MADINAT AL-ZAHRA’ // 07. 2010. PP. 105-122. ISSN: 1139-9996 // ESPACIOS DE VIDA